File size: 3,823 Bytes
0fd3730 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 | """
Test the trained filename anomaly detector.
Usage: python test_model.py
"""
import json, math, os
# --- Load model ---
MODEL_PATH = 'model.json'
assert os.path.exists(MODEL_PATH), f"No model found at {MODEL_PATH}!"
with open(MODEL_PATH) as f:
payload = json.load(f)
hp = payload['hyperparams']
n_embd = hp['n_embd']
n_head = hp['n_head']
n_layer = hp['n_layer']
block_size = hp['block_size']
head_dim = hp['head_dim']
uchars = payload['vocab']
vocab_size = payload['vocab_size']
weights = payload['weights']
BOS = vocab_size - 1
stoi = {ch: i for i, ch in enumerate(uchars)}
print(f"Model loaded: {n_embd}d, {n_head}h, {n_layer}L, vocab={vocab_size}")
# --- Minimal forward pass (float-only, no autograd needed) ---
def linear(x, w):
return [sum(wi * xi for wi, xi in zip(wo, x)) for wo in w]
def rmsnorm(x):
ms = sum(xi * xi for xi in x) / len(x)
scale = (ms + 1e-5) ** -0.5
return [xi * scale for xi in x]
def softmax_floats(logits):
m = max(logits)
exps = [math.exp(v - m) for v in logits]
s = sum(exps)
return [e / s for e in exps]
def gpt_forward(token_id, pos_id, keys, values):
tok_emb = weights['wte'][token_id]
pos_emb = weights['wpe'][pos_id]
x = [t + p for t, p in zip(tok_emb, pos_emb)]
x = rmsnorm(x)
for li in range(n_layer):
x_res = x
x = rmsnorm(x)
q = linear(x, weights[f'layer{li}.attn_wq'])
k = linear(x, weights[f'layer{li}.attn_wk'])
v = linear(x, weights[f'layer{li}.attn_wv'])
keys[li].append(k)
values[li].append(v)
x_attn = []
for h in range(n_head):
hs = h * head_dim
q_h = q[hs:hs+head_dim]
k_h = [ki[hs:hs+head_dim] for ki in keys[li]]
v_h = [vi[hs:hs+head_dim] for vi in values[li]]
attn = [sum(q_h[j]*k_h[t][j] for j in range(head_dim)) / head_dim**0.5 for t in range(len(k_h))]
aw = softmax_floats(attn)
head_out = [sum(aw[t]*v_h[t][j] for t in range(len(v_h))) for j in range(head_dim)]
x_attn.extend(head_out)
x = linear(x_attn, weights[f'layer{li}.attn_wo'])
x = [a + b for a, b in zip(x, x_res)]
x_res = x
x = rmsnorm(x)
x = linear(x, weights[f'layer{li}.mlp_fc1'])
x = [max(0, xi) for xi in x] # relu
x = linear(x, weights[f'layer{li}.mlp_fc2'])
x = [a + b for a, b in zip(x, x_res)]
return linear(x, weights['lm_head'])
def score_filename(name):
"""Return negative log-likelihood (lower = more normal)."""
toks = [BOS] + [stoi[c] for c in name if c in stoi] + [BOS]
if len(toks) > block_size + 1:
toks = toks[:block_size + 1]
keys = [[] for _ in range(n_layer)]
vals = [[] for _ in range(n_layer)]
total_nll = 0.0
for pos in range(len(toks) - 1):
logits = gpt_forward(toks[pos], pos, keys, vals)
probs = softmax_floats(logits)
p = probs[toks[pos + 1]]
total_nll += -math.log(p) if p > 0 else 1e6
return total_nll
# --- Test filenames ---
normal_filenames = [
"acr_banner_spring25_enUS_v01.png",
"acr_email_bf24_enGB_v02.jpg",
"acr_video_demo_enUS_v01.mp4",
"acr_logo_primary_enUS_v03.svg",
"acr_report_fy24q4_enUS_v01.pdf",
]
anomalous_filenames = [
"DELETE_THIS_NOW.exe",
"..hidden_config.bat",
"photo_2024_vacation_IMG_3847.HEIC",
"meeting notes final FINAL v2 (1).docx",
"acr banner spring enUS v01.png", # spaces instead of underscores
]
print("\n--- Normal filenames (should have LOW NLL) ---")
for fn in normal_filenames:
nll = score_filename(fn)
print(f" NLL {nll:7.2f} | {fn}")
print("\n--- Anomalous filenames (should have HIGH NLL) ---")
for fn in anomalous_filenames:
nll = score_filename(fn)
print(f" NLL {nll:7.2f} | {fn}")
|